Claim safety
GroundedNo unsupported numerical, absolute, safety, or environmental wording detected.
Advertising critique and iteration studio
Shape a synthetic product poster, stress-test the message, compare directions, and leave with a brief your team can act on.
Make the direction explicit
Editorial split · Daily commuters
Adjust weights to match the review objective.
No unsupported numerical, absolute, safety, or environmental wording detected.
Practical language and upright product framing support a commuter reading.
Export exact copy, scores, crop notes, claim review, comparison, and next steps.
A strong poster does not ask one beautiful image to solve every problem. It gives product, promise, proof, and action a deliberate order.
Review the edges before the campaign does
Change the delivery format, position the focal point, and simulate common viewing constraints without losing the active direction.
Mechanical checks, human review required
Passes WCAG AA for normal text in the current color pair.
Keep the decisions, not just the files
Confirm what lands first, what earns the second look, and whether the action remains visible after a fast scan.
Judge line length, contrast, density, and language where the poster will actually appear, not only on a large artboard.
Replace absolutes and invented numbers with observable design intent, then route material claims through the right reviewers.
Record the palette, crop logic, tone, hierarchy, and exclusions so the next execution belongs to the same campaign.
One direction. Shared reasons.
How does Poster Lab score creative art direction, and how are accessibility contrast thresholds calculated?
Poster Lab is an interactive review workbench that combines heuristic design scoring, automated keyword-based claim checks, and color contrast calculations for a synthetic commuter concept poster. Rather than serving as an objective industry standard or legal verification, the 0–100 Direction score is calculated entirely in client-side JavaScript by taking a weighted sum across six subjective criteria: hierarchy, legibility, claim safety, audience fit, crop resilience, and brand consistency. Concurrently, the studio measures luminance contrast between active poster colors to determine compliance with Web Content Accessibility Guidelines (WCAG) 2.1.
All direction scores and claim statuses are heuristic client-side approximations and do not constitute legal compliance, advertising certification, or empirical market validation. The 'Claim safety' tool detects basic keyword flags (such as absolute adjectives or unsourced numbers) but cannot evaluate true product claims. Contrast checks evaluate flat background swatches and do not account for variable imagery behind typography.
When the studio initializes with the default Paper white background (#f1f2ed) and dark text, the Active contrast readout displays '15.8:1'. Clicking the second swatch (Cobalt blue, #0f57d6) automatically updates the poster palette and switches foreground text to white (#ffffff). The client-side luminance formula recomputes the contrast ratio between #0f57d6 and white, changing the active contrast display to '6.3:1' and confirming that it satisfies the WCAG Level AA minimum contrast ratio of 4.5:1 for normal text.
Web Content Accessibility Guidelines (WCAG) 2.1 Success Criterion 1.4.3 (Contrast Minimum) requires a minimum contrast ratio of at least 4.5:1 for normal text and 3:1 for large-scale text to meet Level AA conformance. Web Content Accessibility Guidelines (WCAG) 2.1
The contrast ratio is calculated using relative luminance values normalized between 0 and 1 via the standard formula (L1 + 0.05) / (L2 + 0.05), where L1 represents the lighter color and L2 represents the darker color. Web Content Accessibility Guidelines (WCAG) 2.1
Poster Lab's critique matrix scores each dimension from 0 to 100 based on internal rules: copy character count, headline slider scaling, selected aspect ratio, and keyword matching. Each criterion's score is multiplied by its adjustable weight slider, and the normalized sum produces the displayed Direction Score.
Defines the relative luminance calculation formula and the 4.5:1 Level AA minimum contrast threshold for standard text.
The saved Poster Lab HTML separates headline, supporting copy and call to action into three fields. Their native maximum lengths are fifty four, one hundred fifty and twenty eight characters. At two pixels per character, the input limits appear as one hundred eight, three hundred and fifty six pixels. A shorter limit constrains space, not message quality. These limits apply to normal browser text entry; they do not establish that the wording is factual or legible. The saved initial direction score eighty two and contrast fifteen point eight are placeholder text until the application code computes or updates them. They must not be treated as fresh verification. The headline scale range runs from forty eight to one hundred, starting at seventy two. Product prominence ranges from forty five to one hundred, starting at seventy four. At three pixels per slider unit, the two initial settings have bars two hundred sixteen and two hundred twenty two pixels. The HTML supplies range values and output labels separately. Changing a native range does not by itself guarantee that the poster or displayed output text updates. That connection belongs to the referenced application script. The critique list and comparison selectors are initially empty containers intended for that script to populate. This offline source copy contains only the HTML. It references a local application script and visual dependencies which are not present in the saved directory. As a result the native fields can accept text, but the scoring, poster rendering, save variant and export handlers cannot be verified here. Four script references are present; zero corresponding saved script files are available. At sixty pixels per script, the reference bar is two hundred forty and the saved script bar zero. This video therefore does not invent the weighted score formula or claim a working export. It explains the input surface and records the missing evidence. A complete rollout still requires the actual source dependencies and native original versus proposed behavior checks.